Dan: Hello.

Okay.

And welcome back to the next
episode of the quantum divide.

This time round.

I'm really excited to have Can Knaut
from Harvard university joined me.

Can is a.

. At doctoral researcher.

Working at the Harvard university
in the lab of Michel Lukin who

is well-known for his work on.

Neutral atoms.

, What I'm here talking about
with Jan is super exciting.

Is about a paper that's been
released as a result of his

PhD and a huge amount of work.

At Harvard.

University along with Amazon web services.

What they've done is.

entangled two non-local.

qubits Using just a single
photon instead of multiple

photons and entanglement swapping.

And there's also quite a unique.

Method that they've used for
storing the entanglement.

And providing some form of memory.

So this is a really exciting experiment.

And I'm keen to walk through it.

And looking forward to the discussion.

Thanks.

.
Okay.

Well, Thank you, Can, for joining me.

Very much looking forward
to this conversation.

Let's start as I always
do with your background.

Tell us a bit about how you got to where
you are now and we'll go from there.

Can: Yeah, thanks for having me, Dan.

I am currently a graduate student in the
lab of Mikhail Lukin, finishing up my PhD.

And my Path to working on the quantum
network experiments I'm currently working

on was actually not a very direct one.

I initially, after finishing high
school studied business and economics in

Switzerland and finished that and started
working in real estate in Zurich, which

is also host to ETH, which is a very
strong science university and during

my time in Zurich, I was passing the
buildings of ETH and always wondered how

it would be to take these subjects of
physics and math on a university level.

I was always interested in it.

In high school, but up to this
point, I haven't really explored it.

The university level.

And at some point I decided that there's
only one way to find out if this is for me

and I enrolled in a bachelor's of physics
and that was 10 years ago and turned

out to be luckily a very good decision.

During my time at ETH, I.

Could slowly get into conducting
some research on my own.

I started to work with groups working on
quantum dots which are quantum emitters

that, that can be used for certain
quantum, quantum information tasks.

Um, And then I could work for a bit in a
lab working on superconducting circuits.

So I was slowly drifting into the
direction of quantum information science.

And what I liked about that field is first
it had this quite solid and fascinating

grounding in quantum physics, which
is, which is a field theoretically I

really enjoyed learning about, but also
especially experimental quantum physics,

which is the field I got into is a.

It's a very rewarding field
where you also have to do lots

of hands on work in the lab.

You have to actually build devices.

You have to set up labs.

And I slowly over the time realized
that this is an environment I'm

having a lot of fun working on.

And I also realized that the field
of quantum information science is.

It's a very exciting and growing field.

And I decided to pursue my PhD for which
I came to the US to work with Mikhail

Lukin on building quantum networks
using diamond nanophotonic systems.

Dan: Yes.

That's a fascinating journey.

It's always super interesting
when people have come from other

backgrounds, or perhaps gone out of
physics and then come back into it.

Because you get a much more
balanced view, I think different

perspectives of the industry, but
let me ask, first of all what's it

like to work with Mikhail Lukin?

I uh, actually listened to a podcast
interview that he had fairly recently

seems like a very interesting guy and
obviously he's at the cutting edge

with neutral atom computing and so
on what's it like to work with him

and perhaps give us a, what's a day
in the life of, not that there is,

I'm sure a standard day in the life.

I'm sure every day is
different, what's the lab like?

What's the environment like?

Can: Yeah.

Generally working with Mikhail
Lukin looking in his group, which

is fairly big, I think it's quite
quite exciting and unique experience.

I think what is standing out when
I look at my group I'm currently

working on is the kind of size and
scope, so we are roughly 50 people.

In the group and it's divided not only
between several different experiments

in quantum information science, but also
theorists so Working in that group gives

you the opportunity to be exposed to
various different experimental platforms

but also to get input from various
theorists working on Improving schemes for

improving these platforms and I think That
in itself is a very unique environment,

which I personally enjoyed a lot.

That is, this kind of scope you
have is quite useful, especially

when you're an experimentalist.

So for example, I had numerous occasions
when I was working in the lab and say,

setting up a laser and it didn't work.

And I knew that I could just go out of
my lab and I had Seven doors to knock

on of group members who were working on
similar platforms and similar techniques.

And usually one of those doors, a
team member would be able to help me.

So that has been immensely helpful
from the experimental perspective.

Working with Mikhail Lukin
himself, he can also.

He's bringing in that kind of breadth
of experience since he has worked

for many years over various different
platforms, and he can be very useful

in guiding the team members to, to
directions that maybe the individual team

members would not have found otherwise.

So he's really able to.

Help these kind of big strategic
decisions by using the breadth of

his knowledge and also by helping
connect with members of his team.

So overall I had a fantastic
time with this group.

I'm really excited about what will
come out of that group and the future

as I'm now preparing to wrap up.

As of you, the day in the life as a grad
student in that group, I don't think

it's, it looks much different than the
day in the life of any experimental

physics grad student, but to give you
a rough idea of how how it would look

like for someone working on my project.

I work on a project building
quantum networks using emitters

in nano photonic diamond cavities.

And the team I work in is one of the
experimental subgroups containing

roughly five to seven grad students.

And we work on a very team focused way.

So usually we would start out the day by.

Trickling into the lab and preparing
certain measurement scripts.

We would usually have some sort
of meetings or journal clubs over

the days when we would maybe set
up some measurements, then go to

these meetings or discuss papers in
the meantime, and then come back.

And we would also occasionally
meet with the kind of larger group

to discuss intermediate results.

So I think that's this kind of the
schedule might also look different

depending on the stage of the
experiment here in what I currently

described as the kind of workflow.

When the experiment is running, but there
might also be extensive period of times

when you have all hands on deck in the
lab, actually physically building a laser

path on your optical table or modifying
your dilution refrigerator, because

you need to add certain components.

So it's, it can really.

Change from week to week, I would say.

And this aspect is also one of the fun
aspects for me in experimental physics.

It's fun.

You usually don't get bored.

Dan: It sounds great that you've got all
the different groups around you as well.

And have, and being surrounded by experts.

They can help you out.

So I know at Harvard there's also
the Harvard Quantum Institute.

So what is that and how does that
relate to the work that's going on

in The,

the different groups

Can: the,

the quantum initiative is,

Dan: initiative?

Yes, . Thank you.

Can: Yes, the HQI.

So the Harvard quantum
initiative, it's a good question.

It's a It's an initiative that is
aiming to bring together scientists

and engineers to advance the state of
the art of quantum information science.

So that's, we're definitely, our
efforts are really definitely in

the heart of the HQI For me, what
it has meant specifically for my

PhD is that the HQI is has pioneered
a new graduate program at Harvard.

So there's a graduate program for
quantum science and engineering, which

is dedicated to educating PhD students in
the field of quantum information science.

That's a nice addition towards the
more traditional track of physics PhD

students going into quantum information
science and that has also largely

increased the talent pool we have access
to and we've by then by now worked with

numerous grad students from this program.

So this is one very
concrete outcome of the HQI.

More recently, there has also been
a the opening of a new building

at Harvard, which is dedicated
for hosting labs and offices for

quantum information scientists.

And this kind of outfitting and build
out of the building has also been

has also been managed by the HQI.

So it's overall an initiative to I
think focus resources on the field of

quantum information science and Further
double down on Harvard's efforts to

make leading contributions to the field.

Dan: So you've you've been in
Boston for a few years now.

Detect any American accent
that you've picked up there.

It seems like you managed to.

stick fully with your Swiss accent, right?

Your European accent.

What, are there any, any funny words
that jump out from the local dialect

perhaps that you could tell me about?

Can: It's a good question.

Yeah.

I don't think, I don't think I will
get rid of my German accent ever.

I think that's going to stick with me.

Dan: Yeah.

Can: Let me think there.

First of all, it's a little bit.

The environment you're in when you work in
the Cambridge area is quite different from

working in, for example the Boston area.

It's a very international environment,
so it's really not uncommon to

hear English speaking in some
variation of a foreign accent.

So I think there, nothing particular
sticks out, out to me, but the.

Boston or Massachusetts accent
is certainly once you hear it is

certainly something you don't forget.

It is quite interesting, but actually on
campus, you do not hear that often since

it's such an international environment.

And that's maybe also the reason
why I haven't picked it up yet.

Dan: Yeah.

It makes sense.

Yeah.

sure.

University of that kind of that
kind of level is bound to have

a multicultural environment.

So it's surprising.

So let's go into the physics a little bit.

I'm super keen to hear about the
experiment that you've recently

published in nature and beyond, but
perhaps, yeah, let's start with that.

If you could give me the high level view
of that experiment, and then I think

I want to dive down into some of the
individual components of it to understand

the work you've been doing in that space.

Can: Yeah, sure.

So the latest results coming out of
our lab we're focused on building a

two node quantum network, which is
distributing entanglement across a

deployed fiber and to back up a little
bit, I know that there's a big focus

on quantum networks on your podcast.

Definitely know that.

There has been a lot of talks about
motivating quantum networks, but maybe

I'll briefly take a step back and
parse all of these kinds of words.

Separately.

So a quantum network is similar
to a network we use currently, for

example, the internet to communicate
with the key difference that the

information flow using quantum states.

And that makes a lot of the aspects much
more difficult, much more challenging.

The applications for quantum
networks are also quite, quite

interesting and quite different from
the current networks we're using.

One of the key features of
quantum networks is that they

can provide secure communication.

So it is possible to send information
and code it in quantum states.

Thanks.

And have the receiving party
check whether any person has

intercepted that information.

And this kind of check and this
kind of protection is secured by

the properties of quantum mechanics.

And this is a very powerful property,
which if you could deploy this on

a large scale would be very useful.

And Second point of application,
especially in the current era of small

error corrected quantum computers we're
slowly getting into is that you can use

quantum networks to actually connect
smaller quantum processes to build larger,

higher performing quantum computers.

So these kinds of applications
motivated us to look for a platform

that can be built and can actually
be deployed in the real world.

What I mean by that is that.

Quantum network nodes need to be
connected using optical fibers.

And if you want to build a quantum network
that scales beyond just the laboratory

scale, these optical fibers should ideally
be the same ones we're using now to

communicate over the classical internet.

So those should be the same
fiber type and should operate

at the same wavelength, then.

The classical communication fibers.

And that was basically the task we
set ourselves for this experiment.

We wanted to show that we can build such
a quantum network and operate it in a very

robust way over deploy fiber and the key.

Achievement in our work is that we have
distributed entanglement, which is this

non local quantum state, which is seen
as a resource for a lot Quantum network

applications, and we have distributed
this entanglement between two quantum

network nodes through a 35 kilometer
long deployed fiber that was rounded

from Cambridge to Boston and back.

Which contains a really realistic
approximation of the kind of challenges

you would face when when deploying the
actual quantum network, and we could

demonstrate these These entanglement
generation via this deployed fiber.

And we've also shown that we could
entangle various different types of

qubits within our system with each other.

And that's an, in our eyes, really a
step forward towards showing that these

quantum networks and these systems are
working on, could potentially be scaled

up to to be deployed in the real world.

Dan: Great, thanks.

So I know that in this case the technique
used to entangle the, I'll call them

end nodes it's quite unique, it's
the first time I've heard of it, but

before we get to that, let's talk about
the end stations that were entangled.

I believe in this case you're
using silicon vacancies, right?

If you could yeah, walk me through the
details of what's unique or um, what

was it about these particular systems
you created that made them so stable

or just well suited to this task?

Can: Yeah, sure.

You correctly mentioned, this physical
system we're working with is called

a silicon vacancy center in diamond.

It is made out of a very pure carbon
lattice, which is just diamond with

one carbon atom replaced by a silicon
atom that then together with two

vacancies and a diamond lattice forms.

A so called optical defect, which can
be seen as almost an artificial atom.

And these.

SIV centers contain electrons
which then form energy levels

which we can address and we can use
to store quantum information on.

And most importantly for quantum
networking these SIVs or SIV centers

contain a quantum excited state that
is split by an optical frequency

which means that when we send in
photons at that frequency we can

Interact with the quantum system.

And this interaction between light and
the quantum system is really key for any

type of quantum network operation where
you need photons to carry your quantum

information across long distances.

So the reason why the SAV is so useful
and has been so successful for us

is the fact that it has a certain
structure that makes it less sensitive

to environmental charge noise.

So the way to, to look at this as the
following, you can look at these color

centers as a configuration of electrons
that are in a somewhat minimum energy

configuration and a reasonably stable,
but are also sensitive to the environment.

And.

We ideally want to use these SIV
centers, not just in a bulk piece of

diamond, but we want to incorporate
them into structures that can

enhance the interaction between
incoming light and the SIV center.

And there are various ways to do this.

The method we've chosen that has
been also pioneered with together

with other groups at Harvard, namely
the group of Michael launcher, which

is to use nano photonic cavities.

nano photonic cavity can be seen as
the nano photonic equivalent of two

mirrors that form an optical cavity.

And when you have two mirrors
and you can place a quantum

system inside those two mirrors.

You enhance the chances of having
a photon interact with that quantum

system, which can be seen quite
easily by the fact that once you

inject light in such a nanophotonic
cavity, photon bounces back and forth

multiple times between the two mirrors.

And if you now place a quantum system
between those two mirrors, the photon

also passes that quantum system multiple
times, which increases the chances of

interaction between that quantum system.

So what we now do is we engineer a
structure out of diamond that is basically

mimicking these two mirror segments and we
incorporate silicon vacancy centers into

that we call a nanophotonic cavity, which
in turn enhances the interaction between

photons and that nanophotonic cavity.

And at this point, it is very important
that the silicon vacancy center we

use is not sensitive to environmental
specific charge noise because these

Structures we use are nano fabricated,
which contain a lot of surfaces.

Surfaces can potentially trap charges.

So a emitter that can withstand
the kind of noise added by these

charges is required to actually work
with these nano photonic cavities.

And that's also a key difference
between the silicon vacancy center

and the nitrogen vacancy center,
which is also commonly used in

quantum network experiments.

The nitrogen vacancy center is more
sensitive to these charge noises and

is less suited to be incorporated
into these nano photonic cavities.

So that's really far as the combination
we really like this protection from

charge noise induced instability
and our ability to incorporate our

emitters into nano photonic cavities.

Dan: Great.

In terms of the, you mentioned charge
noise is, what does that look like?

Is it a, a kind of like a stochastic
sporadic type sets of noise, or

is it more like white noise that
you can eliminate once you know?

Can: a, this is a good question.

So the way This charge noise would
be visible in your experiments would

be either via drift in your optical
frequencies, which, if they happen

slow enough, you could try to correct
for, but also often can happen quite

fast, which are hard to detect.

This is the, this would be the dominant,
this would be the dominant error source.

And these drifts can be very
drastic in magnitude, which makes

it impossible to really stabilize.

Uh, these drifts are present almost
always in, in the solid state systems,

because any solid state qubit is strongly
coupled to its lattice environment,

so there necessarily will be certain
types of defects that change either

the optical or the spin properties.

But in order to work in these
nanostructures, it is really important

to at least as at first order suppress
the impact of environmental charges.

Dan: yeah, of course.

Uh, or the extrinsic noise
as I've heard it called.

So I understand that you end
up with a dual qubit system in

the silicon vacancy, and you do
that by injecting some isotopes.

Could you give me a bit
of detail on what that is?

Can: Yeah, sure.

So far I introduced the SiV centre
where the queue of freedom we control

is a, basically a free electron or
not a free electron, but an electron.

Part of the, the.

Compound of the silicon vacancy center.

And this electron is a quantum system
we can control and is a quantum system

that couples strongly to incoming light.

And for these reasons, we call this
electron a communication qubit, because

it is the quantum system we use to
communicate between quantum network nodes.

It's important though, that um, when
you start to run a quantum system,

that Have access to a qubit that
can also store quantum information

for extended periods of time.

And this becomes important for scenarios
when you're too quantum network nodes are

actually separated by large distances.

And in these scenarios,
it is important to.

Generate entanglement and then hold
that entanglement for long periods of

time to allow for any kind of classical
signal travel back between the nodes.

And does classical signal
traveling time grows with the

distance between your two nodes.

So what we ideally want is an additional
type of qubit that can store quantum

information for extended period of times.

And as you correctly mentioned we do have
luckily access to such a additional qubit.

We call.

This qubit, the memory qubit, and
it is a nuclear spin, which we get

access to by implanting a specific
isotope of silicon into our diamond

devices, namely the silicon 29 isotope.

And what that does is it gives us
access to an additional nucleus

spin, which is sitting right at the
center of the color center defect.

And which is also strongly
coupled to the electron.

So we can use the electron.

To mediate any kind of interaction
between light and the nuclear spin,

which is very useful if we want
to actually entangle two nuclear

spins into different network nodes.

Dan: Thanks.

So one thing that always comes to
mind for me is you imagine the nucleus

spin of an atom with electrons.

We know that they don't
go in orbits anymore.

It's more like a cloud of
probability, but In this case,

are they two separate things?

Is the electron in the vacancy
on its own, and then the silicon

nuclear spin is separate, or are
they combined within the same atom?

Can: That's a good question.

So the way I would look at this
first is looking at the electron.

What is, what does the
electron actually look like?

And the way I described it earlier
was somewhat of a simplification.

So when you look at these
color centers, you have.

A silicon atom, which kind of tries
to nestle in a diamond lattice.

And specifically it nestles in between
two vacancy sites, which are just

sites called where you just rip out a
single carbon atom, when you rip out

a single carbon atom, you have bonding
electrons from the surrounding carbon

atoms that are just sitting there.

And when you now place your silicon atom
in there, then those dangling bonds of

the carbon atoms in your lattice together
with the silicon Electrons form basically

a new electronic structure that Forms
certain energy levels and these energy

levels are similarly described quantum
mechanically as you mentioned are not

really like orbitals, but more like
probability distributions with certain

probabilities within certain areas
of your defect And the nucleus itself

actually sits at the center of this.

So in a way it is a similar, in a way,
I think your interpretation would be

correct to look at this as comparing
the distribution of an electron and an

actual atom with respect to a nucleus
in that atom, but this, the atom

we're talking about now is a somewhat
artificial system made out of electrons

shared between the carbon atoms in the
lattice and the silicon ion we implanted.

Dan: Nice, thank you.

Dangling bonds.

I think you said that's a new term for me.

Did I pick that up right?

Can: Yeah, dangling bonds.

It's yeah, it's I'm also not the best
chemist, but the way I always picture it

is, yeah, when you rip out a carbon atom
from a lattice, you just have the bonds

dangling and not knowing what to do.

And then you add, you, you
incorporate A different ion next to

it, which gives access to different
additional valence electrons.

And then the system settles into
a new lowest energy configuration.

And these kind of dangling
bonds electrons are part of it.

But yeah, it is a funny term.

Dan: Yeah, cool.

It helps describe the, you
know, I'm a visual thinker, so

it really helps for me anyway.

So let's go into the networking, right?

That's why we're here.

So you using these photonic cavities,
silicon vacancies interact very

well with photons or flying qubits.

Perhaps you want to describe,
yeah, I mean, this is the thing

that I think is unique about this
experiment as far as I'm concerned.

I haven't heard If anything similar to
this, maybe there is, but if you could

describe the, at a high level, first of
all, the, how the two remote vacancies

are entangled and then perhaps we'll
try and dive down a bit more into the

other factors of that architecture.

Can: Yeah, sure.

yeah you're, you're correct in the
sense that the scheme we're using

for entanglement Is when it comes to
solid state qubits unique, and this is

mainly due to the fact that we do have
access to this cavity coupled system,

which allows us to run schemes slightly
differently as opposed to systems

that do not have this strong coupling
between an emitter and the cavity.

There have been similar experiments done
using atoms in macroscopic cavities,

which use similar schemes that we've done.

So the general approach is not completely
new, but it is a unique setting to

do this in a solid state environment
and the way we can think about

distributing entanglement between two
nodes using our system is the follows.

So we first.

Forget the second note and consider
the scenario when we send in a photon

into one of our quantum network nodes.

So this photon is entering the first
quantum network note, which is made

out of a nanophotonic cavity with an
SIV center inside, so that photon now

can interact very strongly with that
quantum system by the kind of mechanism

I described earlier, and we can indeed
perform certain operations on the.

The electron spin of the SIV, and we
can actually entangle the electron spin

of the SIV with the photon we send in.

And the mode we're working in is
a reflection mode, so whatever

we send in into the cavity,
we try to collect it back.

So that allows us to
basically entangle a photon.

with the electrode spin of the silicon
vacancy center in one of the quantum

network nodes and then have that
photon travel through an optical fiber.

And this capability is basically
the key ingredient to generating

entanglement between two electron
spins in two separated quantum network

nodes, which is the goal of a lot
of quantum networking experiments.

And the way that works is that we
basically first perform this mentioned

entanglement generation between a photon.

And an electron in the first node,
after that, that photon, which is now

entangled with the electron in the first
node is traveling in a optical fiber.

And what we now do is we route that
optical fiber to the second node, which

contains the second silicon vacancy
center in a nano photonic cavity.

And this photon now, which still is
entangled with the electron in node 1,

now interacts with the electron in node 2.

Where it also gets entangled with
the electron now in node two.

And we again work in a reflection
regime, so we again collect that

photon after it has interacted with
both the SRV node A and node B.

And at this point, the photon is
entangled with both of the electrons.

And that is really key because now we Can
measure the photon and we can, depending

on the outcome of this measurement,
we can infer whether our two electrons

are in one of two entangled states.

So it's really the photon that
is mediating the entanglement

between the two electrons.

And this kind of reflection based
entanglement generation is really

facilitated by the fact that we do have.

A cavity system around our emitters.

Dan: Okay.

A bunch of questions on
the tip of my tongue.

First one I think is it, the way
you described it is that you're

sending the photon in to the first.

Silicon vacancy, you're not generating
it from within the vacancy because there

are, I know there are ways to generate
entangled photons through pumping in

particular ways into different materials.

But in this case, you're sending it in.

So perhaps tell us about that.

Where does it come from?

Can: Yeah that's a correct observation.

So there are other schemes which
rely on generating single photons

through the quantum system.

So if I dive a little bit more into
detail, how the interaction between an

incoming photon and a silicon vacancy
center works it's the following.

So we.

Have our silicon vacancy center
coupled to a nano photonic cavity.

And what we can now observe is that
when we send in light at a very

specific frequency, we get it reflected
off that cavity SIV system only if

the SIV is in a specific spin state.

So if say the SIV is in a spin down state,
the light does not get reflected off.

If it's in a spin up state,
the light gets reflected off.

This highly spin dependent
reflection contrast is directly

a product of the strong coupling
between the cavity and the emitter.

And the strong reflectance
contrast is also the underlying

mechanism of our entangling gates.

So intuitively, the way to see this is
that if I have A interaction which is

strongly dependent on the electron spin
state, which is, maybe I forgot to mention

this, which is the degree of freedom we
use to encode our qubit, the spin qubit,

then I can use the strong interaction
to also perform entangling operations.

And the reason why this is not more
commonly used is because indeed you

would need a strong interaction between
your cavity and your quantum system.

Which is the case for certain quantum
systems, for example, atoms and

cavities, but it's, for example, not
the case if you look at NV centers

in bulk diamond, which do not have a
nanophotonic cavity around, which cannot

provide the strong reflection contrast.

So the fact that we use photons, we
prepare initially and then bounce off

our cavity Si V system is a direct.

Consequence out of the fact that we have
this strongly coupled cavity QED system

Dan: Okay, cool.

Spin independent.

I think that should be rebranded as
a new word, just spin dependent, just

Can: spin dependent.

Dan: lot

Can: Yeah.

Dan: Although it is in fact
dependent on the spin, right?

Can: on the spin and yes,
that's very important.

Dan: a result then you, does that
mean only two of the four potential

Bell states can be formed between the
photon and the electron at that point?

Can: Yes.

So our techniques specifically generates
two out of the four possible belt states,

but it's possible to also, then once
you have any belt state, you can just

perform single qubit operations to.

To move them into any
of the four Bell states.

Dan: Yeah.

Yeah.

So the next question
is about the cavities.

I'm assuming that because they have
to interact with the same photon, they

have to be precisely the same size.

are they fixed at a particular
size or are they dynamic anyway?

Can: So with I assume
you mean the frequency

Dan: Yeah,

the gap between the

Can: the, the cavity
resonance frequencies.

Yes, exactly.

So actually we do not use what we
call macroscopic cavities, which

is probably what you're thinking
about when you think of a cavity,

which is actually two mirrors.

These types of cavities would
be used, for example by.

Experiments using atoms

Dan: like an iron trap
or something as well.

Yeah.

Can: Ion traps usually don't use cavities.

There's some experiments that do this.

It would mainly be neutral
atoms trapped into cavities.

And these would indeed use, um, free
space mirrors very close to each other.

What we use is something
a little bit different.

We use something called a
photonic crystal cavity.

So that's basically made out of
a, it's a completely monolithic

structure made out of diamond.

And the, the structure itself can
be can be visualized as a, it's a

triangular cross section beam of diamond.

So I like to describe it as a Toblerone.

All the Toblerone bar, which is a
triangular cross section beam, if you

want, and then what we do in addition to
that uh, we pattern holes into this bar.

And these holes when they're
placed in a certain periodicity can

basically mimic mirror segments.

So with that, we can basically uh,
design and build a monolithic diamond

device, which acts as an optical
cavity and back to your question

in terms of the frequency, the
resonance frequency of these devices.

You're absolutely right that
these resonance frequencies

of the cavities must be.

Close to the SiV transition
frequencies, and that's something

that is simulated beforehand and
then targeted during nanofabrication.

We have additional ways to shift the
frequencies of the cavities inside

our experiments by depositing gas
on it, which changes the reflector

refractive index of our cavities.

And which allows us to slightly nudge
the resonance frequencies of the

cavities by a couple of nanometers.

Dan: Okay, so that's your really
fine grained control, just to

tune the system if you like.

Can: Yeah.

Dan: Okay, so the photon exits,
goes down the fiber, it gets

to the other silicon vacancy.

There's another interaction, and at
this point there are three elements

which are in a entangled state, right?

So Is this a GHZ state at this point?

Can: Yeah, correct.

So after the photon is interactive
with the second SRE you can describe

the quantum state of the photon
and two SiV as a GHZ state indeed.

So if you were to write it down in,
in, in quantum formalism, it would be

The two electrons would be up, and our
photon would be one two electrons

down and the photon being zero.

So it's indeed a GHZ state.

And after measuring out the photon in
a different basis than the zero one

basis, but in this position basis,
we do project our two electrons

into one of the two bell pairs.

So what we end up with is either.

Up, plus down with probability one half or
up, minus down with probability one half.

Dan: Okay, nice, and that's why you
can then directly and deterministically

understand what the entangled
states are of the qubits and the

electrons without observing them,
which then means you can then use

them as in some kind of information
transfer teleportation type process.

Can: Yes, exactly.

So it is an important
feature that our entanglement

generation is indeed heralded.

So we can measure the photon.

And as you correctly said that
we know which state we're in

and we can proceed from there.

Dan: So the question is
this method of heralding.

With, from a GHZ, GHZ state obviously
all the English listeners are going

to be falling off their chair now,
I should say GHZ state, but is it

more efficient or optimal in any
way than heralding with two photons

received from two different sources?

For example, if you have two ion
traps, both emitting a photon and then.

Interfering them together in
the network somewhere, in a

Bell state measurement device.

Can: That's a good question.

So those are generally two different
techniques to generate entanglement.

And I think one of the big differences
between the techniques is the need for.

Phase stability between two links.

So what you described as a method to
entangle, for example, trapped ion systems

or NV centers outside of cavities is
where you basically use an emission based

scheme on two systems, which both emit
single photons that then travel through.

Two separate paths and interfere
with each other on a beam splitter,

which is usually placed in the middle
between the two quantum network nodes.

Dan: Yeah, that's the example I was

thinking of.

Can: And these schemes they work really
well and have been used for various

different platforms very successfully.

But one of the downsides of these
schemes is that the two photons

each travel different paths.

That means that.

These two paths need to be stable with
respect to each other, specifically

that the phase each photon acquires
through this path should be ideally.

Be the same.

And that's something that is definitely
easily achievable or not easily,

but it is achievable through locking
techniques in the lab environment,

and it's a bit more challenging to do
so in a deployed environment and our

scheme evolves around the fact that
all the photons travel the same path.

Which means that the requirements on
phase stability are much, much lower.

So when comparing the two schemes this
is definitely one of the big differences.

I would say where our scheme stands out
when it comes to efficiency it depends

on the system you're actually using.

And there are certain configurations
where a scheme with beam

splitter in the middle be more
efficient because certain losses.

Don't enter more than once, for
example, any kind of photon extraction.

Loss from any of the nodes on interest
once in such a would call them herald

scheme as opposed to scheme we are using,
but at the end it will boil down to the

actual realized losses in the systems.

So there are definitely differences
in rates that can be achieved

between the systems, but.

What that actually means in
terms of entanglement rates and

hertz really depend on the actual
physical system you're working with.

Dan: Yes, exactly.

it always depends.

That's always the right answer.

But But of course in, in these
complex systems there's so many

different variables at play.

But that's interesting to know that,
the fact that there's a shared path, a

single path, is significant actually.

Can: Yeah, it makes a difference
and it was quite relevant for our

experiment where we did indeed
generate entanglement through a 35

km long fiber link which was routed
through a very busy urban environment.

This kind of task with a parallel
entangling technique would have

been much more challenging because
it would have involved actual phase

locking of very noisy fiber paths.

And we did not have to
luckily deal with that.

We only had to perform
comparatively simple polarization

stabilization through that link.

So we could immediately benefit
from our architectural choice

by generating this entanglement
through this noisy fiber link.

Dan: Okay, nice.

So what about the rates?

You did mention them.

I know that obviously you're building
an experiment here, not a production

system, but what kind of rates were
you getting and how do you think

they could be optimized or improved?

Can: So, rates we would be getting, if you
look at electron entanglement generation,

the faster that we can go is roughly one
Hertz, which in the realm of entanglement

generation is least order of magnitude
wise, close to the state of the art.

It's not the fastest that have been done,
but it is definitely a respectable number.

Of course, when we look
at actual required.

Rates for any useful, for example,
quantum communication, ideally would

like to have these numbers increased
by a couple orders of magnitude.

But as you mentioned, this is first
proof of principle demonstration.

And there are definitely.

A lot of avenues we, we can go
down to improving these rates.

So for example, just by working with
emitters that are slightly better behaved,

that don't need to be reset as slowly
as the ones we were working with you

could already increase the repetition
rates of our experiments significantly,

which is one way to improve on rates.

There are also various improvements on.

Coupling fibers into nanophotonic
cavities that has been made over the

last two years, which can, could be
incorporated in these experiments,

which would further increase the losses.

So they're definitely already very
straightforward ways to improve these

rates and they're also more forward
looking approaches that would involve

setting up multiplexed experiments
where you would have say hundreds

of these nanophotonic devices.

Which are actively switched to
significantly boost entanglement rates.

So that's I would say a mixture
between very short term improvements.

We were working on that could
increase these rates and more

architectural improvements that also
the larger community is working on.

That should be able to move these
rates into a regime where it

becomes relevant for applications.

Dan: Yeah, I mean, that's the beauty,
I think, of this type of technology.

It's on such a small kind of angstrom
scale that if there can be some

multiplexing of some kind, it's not
really going to impact the the system.

size to make any bulkier.

It may make the control, a little
bit more difficult to manage, but

Can: it is a non trivial problem
to, to multiplex these systems.

It is not as easy as just Putting
a hundred of these in a row, this

is, it's a significant research
effort to design cavities and design

devices in a way that is conducive
to these multiplex experiments.

But it's definitely something that
is that is, that, that should be

doable and with enough efforts
put in should also materialize.

Dan: yeah, it's good to see these
avenues of research, which I think

ultimately would be taken up by
other people, or maybe Maybe even

in, in the lab where you're working.

So that brings me to the system
as a whole, we're talking about

multiple silicon vacancies, the
fiber in between it, the control

systems, the monitoring systems.

What would you say?

Yeah, sticking to the
point we were on there.

Are there students in the.

Cohort there, which are a few years
behind you that are taking on some of

these scaling issues or some of these
tasks to optimize the processes or

industrialize the control, or, what
improvements will be on the agenda

for Mikhail Lukin's lab and the wider
team as a result of this experiment.

Can: Yeah, absolutely.

So my.

My team, which is roughly
seven grad students is still.

Very active in the field.

We are actually expanding
the number of nodes to three.

So we have a third quantum network
note built up and we're currently

starting to run experiments on this
third quantum network note in addition

to the two we've been used so far.

So there's definitely a lot of activity
centered around trying to figure out.

What we can do with this newly
gained ability to entangle

out our quantum network nodes.

And how to improve on this, how
to improve the number of qubits we

have access to, how to improve the.

The fidelities of the control and also
how to scale us up to, to experiment

running on, on three nodes, which is
interesting because when you have three

nodes, you could actually start thinking
about building a quantum repeater

experiment, which is, would be a crucial
ingredient to scaling up quantum networks.

And that's definitely something
where we're thinking about and where.

Also junior team members in my team, I'm
sure we'll make significant contributions

to in the next couple of years.

Dan: Cool.

I think I was going to ask with three
nodes, do you end up with another

entangled state in the same system,
but because you said repeaters, that's

making me think that actually each
individual pair is entangled together.

So you end up with a potential kind of
triangle of connections, or you have one

in the center and vacancy in the center
is, is acting as a repeater because it's

connected to the other two on like a.

An arm on the left and an
arm on the right type thing.

Can: Yeah.

So we haven't started these three
node experiments yet but as you

mentioned, there are various different
topologies you could think about.

The kind of typical proposal for a quantum
repeater architecture is the one which you

mentioned where you would have one node in
the middle, which would generate pairwise

entanglement with the two outer nodes.

And then would use a entanglement swapping
technique to generate entanglement

between the two outmost nodes.

But there's definitely other interesting
network topologies we could think

about both in the context of quantum
repeaters or in, in other contexts.

Dan: Okay, cool.

I'll look out for those.

So yeah, in terms of the system as a
whole, did you have any partners you

wanted to call out or collaborators that
were working with you through this period?

Can: Yeah, these absolutely.

These quantum network
efforts are it's a lot of.

Teamwork involved that not only
requires a fairly big team of batches.

students working in the lab, but
also I would say bigger ecosystem of

collaborators, namely other groups of
Harvard, for example, a group of Marco

Lonchar, who has really pioneered the
diamond nano fabrication necessary to

work with our nano photonic cavities,
plus a collaboration with the center

for quantum networks, which has also
been very useful for this experiment.

This the collaboration with AWS came
to light via a couple of graduate

students from my experiment who joined
AWS a couple of years ago and who

are now working on tackling similar
problems than we in an industry setting.

And they have been very useful in
providing support in nano fabrication as

well as giving scientific input on the
experiments we're currently involved in.

Dan: Nice.

Thank you.

let's move on to moving
towards wrapping up.

This has been a fascinating conversation.

even having kind of read the abstract
and had my first skim through a

couple of these the the paper in
nature and, and other articles.

I still have loads of questions on
how this type of system is working

at a low level, but it's really
exciting to talk to you about it.

But yeah, so I like to
ask a few questions.

So, what would you say is your
favorite do you have a favorite paper

or favorite influential piece of work
in the quantum domain that really

was particularly influential for you?

Can: That's a tough one there.

There are a lot of very
important papers and they're also

luckily growing by the minute.

But if I had to look back at papers
that really Shape the field of

quantum networking and maybe even
quantum information overall, I would

have to point to a paper by Charles
Bennett and Jill Psar, which are.

The two researchers that invented the
first quantum cryptography protocol, which

is named BB 84 BB for the two last names.

And then 84, because that's
the year they invented it.

And I think that the paper is
called something like crump to

quantum cryptography and then
some, some additional title.

And the interesting thing about this
paper is that it really was one of

the first works that have shown a real
quantum advantage over classical methods.

So this paper proposed basically a scheme
to use photons that encode states in

the polarization basis and demonstrated
a technique where you can use these

photons to distribute a cryptographic key.

And have it secured from any kind of
eavesdropping in a way that is information

theoretically secure, which means that
the receiving party can perform certain

measurements and it can prove that no one
else could have possibly have access to

enough information to recover the key.

And this kind of realization
which I think also at the time

was not necessarily fully.

Acknowledged by the physics community.

I think this paper first was not published
in the physics journal but a electrical

engineering journal was yeah, one of the
first times when researchers actually

found a application for quantum physics.

Where this application is clearly
outperforming any classically

possible operating system.

And this has been hugely
influential for developing the.

The field of quantum networks, obviously,
because quantum communication is one

important aspect, but I think it has
also motivated uh, researchers to look

more into more complex quantum circuits
that, and quantum algorithms that can be

produced to produce a quantum advantage.

So this is really, I think if I had to
pick one paper, which I really like,

I think it's going to be that one.

Dan: That's a good choice.

Only 40 years old.

Can: Yeah.

Dan: Which is uh, I mean, you
know, there are production systems

in the market now using BBB 84.

And and that's super interesting.

And that's even after there
was what B 92 and e er in 1991.

These protocols are deployed in systems
in production now just 40 years later.

But that's great.

Thank you.

Thanks for raising that.

Give us a bit of a, your vision for the
future of quantum networking, quantum

computing, is there a particular know, you
get a lot of kind of fluff in the media

about you know, and hype about where we're
going, but from your perspective and how

close to the work on the ground you are,
realistically speaking with, is there

some kind of scenario you could describe
that would be interesting to hear looking

forward into your, into your crystal ball?

Can: I wish I had such a crystal ball
and I think I could make a lot of money.

No I, I'm obviously closest to
developments in quantum networks.

And I think in that field, we've
definitely seen quite tremendous

progress over the last couple of years.

And I do anticipate that
progress to continue.

I think what we'll see in the
field of quantum metrics is

probably somewhat of a shift of.

Networks that are more focused
towards quantum communication and

networks that are more focused
towards distributed quantum computing.

And the latter part will probably co
evolve also with the kind of buildup of

larger quantum computers, which in itself
is also a field that is, that's something

I'm obviously indirectly exposed to.

We have my presence in the larger.

Harvard and MIT quantum information
community, but I have less like personal

insight in if I had to make a guess how
the field of quantum computing would

involve is I would probably assume that
the pace of the last one or two years,

which in my opinion has been picking
up will hopefully accelerate for the

next couple of years towards Having
more problems focused around actual

logical qubits and how to start running
algorithms on logical qubits and not

physical qubits anymore in terms of.

of systems and usefulness of calculations.

I think this is a very
hard question to answer.

I do think there's going to be a
lot of interesting developments

in the next couple of years in
the field of quantum computing.

And definitely be excited to, to follow
these or be part of these demonstrations,

but it is really hard to make a definite
statement about when say the first

useful quantum computer will exist.

Dan: Oh, Hey, mean, I kind of avoided
that question, but it's what some people

call it, the chat GPT moment, right?

yeah, thanks for that.

I think I'm seeing the same kind
of thing where there's a bit more

of a focus on fidelity and moving
towards fault tolerant computers than

just scaling intermediate systems
which are noisy as much as possible.

So yeah, it's going to be interesting
to see what part networking,

quantum networking specifically
plays in the scaling that will

come during and after that.

Can: Yeah maybe one thing to
mention is that I do think that

on the longterm horizon, there's
a very strong use case for using.

Even small quantum networks for
distributed quantum computing.

And the kind of logic there is that
any quantum computing system is trying

to increase the number of qubit sizes
per single computing unit and that

computing unit can be like a dilution
refrigerator full of superconducting.

Chips, for example, or it can be
a vacuum chamber full of trapped

Atoms, but whatever the system, there
will necessarily be a point where

the container for your qubits will
not be able to grow any further.

And at this point the only real
way to increase the computational

power of your quantum computers is
to start to link them up together.

And at this stage, you will
necessarily need a quantum network.

You can also call it a quantum
interconnect, but there's definitely

On the long term vision, the
need to link quantum computers

and quantum process together.

So I do think that especially for the
field of quantum networks, the kind of

development of quantum computing is very
relevant, and I hope that in a couple of

years, these, some of these developments
will go hand in hand to actually start

improve the computational power of quantum
computers by linking them together.

But it's also very.

Technically challenging problem.

So I do not anticipate this happening
overnight, but I'll, I'm very

excited to, to keep my eyes out
for any developments in that field.

Dan: Yeah, no doubt there'll be many.

I'm absolutely aligned with you on
the need for networking when it comes

to distributing quantum computers
to maintain shared entanglement

across a distributed system.

It's it's really exciting
to be part of and watch.

I Am going to wrap up now, but I'm
going to wrap up with a final question.

stepping away from the science for a
minute, Can what do you do to wind down?

What do you do outside of physics
that perhaps you could share?

Can: Yeah, sure.

I think it really has changed over the
years, but one thing that, that I try to

consistently do during my time here is
to have some sort of physical activities.

So could be just walking
out or going for a run.

Together with that, the New England area
around Boston is actually quite beautiful.

You have a lot of different options to
either go hiking or go to the beach.

And I definitely when time
permits try to make use of that.

And then I'd say for the last couple
of years, one, one hobby I also picked

up, which is also Involved a lot of
other physicists in my team and other

friends in the game is playing poker.

I really like to play poker.

So I would have my poker rounds at
my place and would invite people

over and it would sometimes be half
physicists, half non physicists.

But usually we would try to wind down
not talk too much about work at physics

and just have a good time together.

And those have been.

Really nice to wind
down and also socialize.

Dan: Sounds great.

Yeah, I didn't know about the
kind of area around Boston.

What's the temperature
like on the beach there?

Is it too far north to
be enjoyable very often?

Can: Wouldn't necessarily
go into the water.

It is pretty chilly.

I'm not sure it's probably,
it's definitely below 20 C.

I don't know.

I don't know exactly how cold
it is, but it is pretty cold.

So when I would go to the beach,
it would be more to just hang

out at the beach and read a book.

Then to go into the water, but you can
and you can actually surf there too.

There are people who are Who
like to surf around cape cod?

So it all depends on how
brave you are, I guess

Dan: yeah.

And So is that is that Texas Hold'em?

Is that one of your,
is that your favorite?

I think that's the one that definitely
needs money involved, isn't it?

Can: There's always money involved,

Dan: Okay.

Yeah.

Can: and not large amounts we usually
mainly play for fun, but You need to

have some amount of money involved
to force people to make Yes, we would

play Texas Hold'em, and there are other
variations of it we would occasionally

mix in to keep it interesting,
but it's mainly Texas Hold'em.

Dan: That's great.

Thanks.

That's uh, it's fun to hear.

Cool.

Okay.

Well, We'll wrap it up then.

Thank you very much again.

I really enjoyed this talk and I'm
keen to keep an eye on what, what's

next for you and further progress
in the group under Mikhail Lukin.

So thank you very much for the overview.

Bye.

Can: Yeah, thanks so much for having me.

It's been a blast.

Dan: I'd like to take this moment to
thank you for listening to the podcast.

Quantum networking is such a broad domain
especially considering the breadth of

quantum physics and quantum computing all
as an undercurrent easily to get sucked

into So much is still in the research
realm which can make it really tough for

a curious IT guy to know where to start.

So hit subscribe or follow me on your
podcast platform and I'll do my best

to bring you more prevalent topics
in the world of quantum networking.

Spread the word.

It would really help us out